The vagus nerve mediates the physiological but not pharmacological effects of PYY_3-36 on food intake

Abstract

Peptide YY (PYY_3-36) is a post-prandially released gut hormone with potent appetite-reducing activity, the mechanism of action of which is not fully understood. Unravelling how this system physiologically regulates food intake may help unlock its therapeutic potential, whilst minimising unwanted effects. Here we demonstrate that germline and post-natal targeted knockdown of the PYY_3-36 preferring receptor (neuropeptide Y (NPY) Y2 receptor (Y2R)) in the afferent vagus nerve is required for the appetite inhibitory effects of physiologically-released PYY_3-36, but not peripherally administered pharmacological doses. Post-natal knockdown of the Y2R results in a transient body weight phenotype that is not evident in the germline model. Loss of vagal Y2R signalling also results in altered meal patterning associated with accelerated gastric emptying. These results are important for the design of PYY-based anti-obesity agents.

Highlights

• •A quarter of vagal afferent neurones express the NPY Y2 receptor (Y2R).
• •Loss of vagal afferent Y2R abrogates the anorectic effects of low dose but not high dose PYY_3-36.
• •Loss of vagal Y2R accelerates gastric emptying and alters meal patterning, leading to smaller, faster and more frequent meals.

The anorectic gut hormone peptide YY (PYY) is secreted from enteroendocrine L-cells (EECs) in response to food intake in a manner dependent on both caloric load and macronutrient composition [[1], [2], [3]]. Peripheral administration of the active circulating form, PYY_3-36, powerfully reduces food intake in rodents and humans [4], and chronic administration induces weight loss in rats^, [5]. PYY_3-36 retains its anorectic effects in obese subjects [6], suggesting potential utility as an anti-obesity pharmacotherapy. However, the mechanism of action of this peptide is not fully understood. Given the challenges of maintaining the effects of PYY_3-36 chronically, and potential side effects including nausea, understanding how this system physiologically regulates food intake may help unlock its therapeutic potential whilst minimising unwanted effects.

PYY_3-36 mediates its anorectic effects via the neuropeptide Y (NPY) Y2 receptor (Y2R) [4], but the site of action is unclear. Pharmacological administration studies have reported direct effects of PYY_3-36 on the brain [4,7]. It has been postulated that circulating PYY_3-36 crosses the blood–brain barrier and accesses Y2Rs in the hypothalamic arcuate nucleus (ARC) to inhibit feeding [4,7,8]. Here, Y2R acts as an auto-inhibitory presynaptic receptor in arcuate NPY neurons. PYY therefore inhibits NPY neurons resulting in the loss of tonic inhibition of neighbouring anorexigenic proopiomelanocortin (POMC) neurons [4].

However, other studies suggest that PYY_3-36 does not act directly on the hypothalamus to suppress food intake [9,10], and PYY_3-36 retains its anorectic effect in POMC-deficient and melanocortin-4 receptor (MC4R)-deficient mice [11,12], suggesting that the hypothalamic melanocortin system is not essential for the anorectic actions of PYY_3-36. The vagus nerve, the main neural link between the gut and the brain, also expresses the Y2R in rodents [9,13] and in humans [13]. Furthermore, Y2R expression levels in the vagus are regulated by nutritional status [13]. The role of vagal afferent signalling pathways in the gastric emptying or anorectic effects of gut hormones remain incompletely understood. Some rodent studies suggest that the anorectic effects of PYY_3-36, including upstream signalling to the hypothalamus, are abolished with subdiaphragmatic vagotomy in rodents [9,14]. Others present evidence that central anorectic effects are preserved after surgical vagotomy [10] or the ablation of sensory afferents using systemically-administered capsaicin [15]. However, surgical and chemical vagotomy lacks fibre or receptor specificity, which can result in confounding effects on appetite and gastrointestinal function that make it difficult to determine which of the physiological effects of PYY_3-36 are mediated via the vagus.

We hypothesised that the physiological anorectic effects of PYY_3-36 are mediated by the vagus nerve. Here we demonstrate that animals bred with reduced NPY Y2R expression in sensory neurons (Nav1.8/Y2R KO), including the afferent vagus, would demonstrate an attenuated response to low but not high dose PYY_3-36. We also generated a mouse model of adult Y2R knockdown specific to the afferent vagus. We bilaterally injected an adeno-associated virus expressing cre recombinase (AAV-Cre) into the nodose ganglia (NG), extracranial structures that contain the cell bodies of vagal afferents, to achieve selective knockdown of the Y2R in the vagal sensory neurones of mice with a floxed Y2R (NG Y2R KD). We demonstrate that postprandially-released PYY_3-36 requires intact vagal signalling but that pharmacological doses suppress appetite despite loss of vagal afferent Y2R.

Nav1.8 is a sodium channel isoform expressed in sensory neurons of the vagus nerve (over 75% of mouse vagal neurones) as well as in the spinal cord [16]. Y2R^loxP mice were bred with mice expressing cre recombinase driven by the Nav1.8 promoter to generate Nav1.8/Y2R KO animals. In this germline model, Y2R mRNA expression as measured with qPCR was significantly reduced by 60% in both the left and right NG of Nav1.8/Y2R KO compared to littermate controls (Figure 1A). No differences in adult body weight or body composition were noted between Nav1.8/Y2R KO and control mice (Figure 1B).

Figure 1.

A. Relative expression of Y2R mRNA in left and right NG of Nav1.8Cre/Y2RloxP/loxP mice (Nav1.8/Y2R KO) and littermate controls that were sacrificed at 16–30 weeks of age (n = 4–5 mice). Expression was normalised to the Y2R expression in the right nodose ganglion (NG) of control mice. ∗∗p < 0.0001 by two-way ANOVA followed by Bonferroni's test. B. No differences in adult body weight between Nav1.8/Y2R KO and control mice. C. Low dose PYY_3-36 suppresses food intake over 2 h post-injection in littermate controls but not in Nav1.8/Y2R KO mice. In contrast, high dose PYY_3-36 suppressed food intake in both groups. D. Relative expression of Y2R mRNA in left and right NG of Y2RloxP/loxP mice injected with AAV-Cre into the NG (NG Y2R KD) and of littermates injected with control virus (AAV-GFP) into the NG (NG Y2R +) that were sacrificed at 17–21 weeks post-surgery (n = 13–14 mice). Expression was normalised to the Y2R expression in the right NG of control mice. ∗p < 0.0001 by two-way ANOVA. E. Exemplar image of NG cross section from control animal (NPY2R floxed injected with AAV-GFP). Cell morphology, highlighted by haematoxylin staining was used to identify neurons; magenta staining marks GFP staining and turquoise for Y2R. F. NG cross section from NG-specific NPY2R knockdown animal (NPY2R floxed injected with AV-cre). Haematoxylin staining identified neuronal nuclei, magenta staining marks Cre staining and turquoise for Y2R. G. Early body weight phenotype in NG Y2R KD mice in the first 4 weeks following viral induction of cre mediated Y2R KD. H. Food intake in NG Y2R + and NG Y2R KD mice over 6 days at week 7 post-surgery.

We investigated whether the response to exogenous PYY_3-36 was dependent on vagal Y2R signalling. The effects of intraperitoneal administration of low and high dose PYY_3-36 was investigated in Nav1.8/Y2R KO animals compared with littermate controls. The low dose (3 μg/kg) had previously been shown not to cause conditioned taste aversion (CTA) [10] and was compared to a high dose (30 μg/kg) more likely to access centrally expressed Y2R [4,10,12]. Here, neither dose caused a conditioned taste aversion response in wildtype mice (Supp Fig. 1).

Interestingly, low dose PYY_3-36 suppressed food intake in littermate controls but not in Nav1.8/Y2R KO mice. In contrast, high dose PYY_3-36 suppressed food intake in both groups (Figure 1C). These results suggest that low-dose exogenous PYY_3-36, which may better reflect the actions of endogenous PYY_3-36, acts via Y2R vagal pathways, but that high-dose PYY_3-36 bypasses vagal signalling.

To confirm that these findings did not reflect developmental differences, we generated a vagus-specific, adult knockdown model by administering bilateral intra-NG injections of AAV-Cre in Y2R^loxP/loxP mice on a C57/Bl6J background (NG Y2R KD). This resulted in a significant bilateral reduction in Y2R mRNA in the NG, as measured with qPCR at 20–24 weeks post injection, compared to controls injected with a control adeno-associated virus expressing green fluorescent protein, AAV-GFP (NG Y2R +) (Figure 1D). To further characterise the nature of the knockdown achieved with this approach, we undertook RNAscope to quantify Y2R expression in the NG Y2R KD versus the NG Y2R + animals. In the control group, of the 2824 neurons examined, 623 stained positively for Y2R mRNA (22%). Of the 4149 neurons examined in the KD group, 61 stained for Y2R (1.5%, p < 0.001 on chi-squared comparison) (Figure 1E–F). Of note, the proportion of neurones staining positively for GFP in the control group and cre in the knockdown group were 26% and 28% respectively.

Animals were group-housed in the postoperative period for welfare reasons, precluding individual food intake measurements. Body weight was significantly higher in the NG Y2R KD group up to 6 weeks following NG injection, when the body weight curves converged (Figure 1G). By week 7 post injection, body weight and food intake in each group (NG Y2R KD vs NG Y2R +) were similar (Figure 1H).

To confirm that the Y2R knockdown was specific and that other vagal afferent signalling pathways remained intact following NG injection, we investigated the response to cholecystokinin octapeptide (CCK-8), a gut peptide established to exert anorectic effects via vagal afferents [17]. CCK-8 at 5 μg/kg (intraperitoneal, IP) significantly reduced subsequent food intake compared to vehicle in both groups (treatment component on two-way ANOVA after 1hr, p = 0.0044) and no difference (p > 0.9 on ANOVA with multiple comparison correction) was observed in CCK-8-mediated food intake reduction between NG Y2R KD and NG Y2R + mice at 1 h (Figure 2A). The abrogation of the feeding effect of low but not high dose PYY_3-36 was recapitulated in this adult selective knockdown model of vagal afferent Y2R (p = 0.033 on ANOVA with multiple comparison correction, Figure 2B). To directly confirm a functional phenotype, calcium imaging of cultured NG neurones revealed that 35/130 (27%) neurones from the control injected animals whilst only 9/116 (8%) neurones from the cre-injected animals were responsive to the addition of PYY_3-36 (p = 0.001 on Х² test). Of those neurones that did respond to PYY there were no discernible differences in calcium oscillatory activity between NG Y2R KD and NG Y2R + (Supp Fig. 2).

Figure 2.

A. Cumulative food intake (FI) after vehicle or CCK-8 (5 μg/kg, IP) administration in the early dark phase (n = 13–14). ∗p < 0.05 (time x treatment) by three-way ANOVA. Two-way ANOVA followed by Dunnett's test to compare both time 1 h and 2 h to time 0 h resulted in p < 0.0001 for both time points in both control and NG Y2R KD. B. One-hour FI after vehicle or PYY_3-36 (3 μg/kg or 30 μg/kg, IP) administration in the early dark-phase in (AAV-Cre injected) NG Y2R KD vs littermate control animals (n = 11–15). ∗p < 0.05 by two-way ANOVA followed by Sidek's multiple comparisons. C. The feeding response to a gavage of either a high calorie meal or an equal volume of water was compared in NG Y2R KD and NG Y2R + control (AAV-GFP injected) mice. Cumulative food intake (FI) over 2 h after gavage of water is shown in on the dashed lines, with no differences observed between the NG Y2R KD (orange) and control groups. However, after gavage of a caloric meal, cumulative 2 h FI in the early dark phase was significantly greater in the NG Y2R KD group in line with the hypothesis that vagal Y2R mediates the anorectic effects of postprandially release PYY (n = 12–13). ∗∗∗∗p < 0.0001 (time x treatment) by three-way ANOVA. ∗p < 0.05, ∗∗p < 0.01 by two-way ANOVA followed by Bonferroni's test. Two-way ANOVA followed by Dunnett's test to compare to time 0 (at t = 1 h in Bolus condition, p < 0.01 for both groups; at t = 1 h in the control condition, p < 0.001 for NG Y2R + and p < 0.0001 for NG Y2R KD; at t = 2 h, p < 0.0001 in all conditions and groups). D. Plasma acetaminophen levels after oral gavage of a 20% glucose bolus containing 1% acetaminophen in 9-h fasted animals (n = 7–3). ∗p < 0.05 by multiple t test with Bonferroni's test. E. CLAMS data averaged between phenotypes over all 10 days (n = 5–6 per group) for average meal weight, meal duration and intermeal interval (all events included).

To study the role of the vagal Y2R in response to physiological PYY release using this postnatal knockdown model, postprandial food intake was measured following oral gavage of a mixed nutrient meal (5.3 kcal/mL, energy: 22% CHO, 69% Fat, 9% Protein). This meal was confirmed to induce a more than six-fold rise in total plasma PYY levels in wild-type C57 mice, from 10.10 ± 9.21 pmol/L to 69.95 ± 21.60 pmol/L at 30 min post gavage (n = 4–5; p = 0.0534) (Supp Fig. 3a). This rise in PYY is similar to that previously reported in mice and in humans postprandially [1]. Ad libitum food intake was significantly reduced at 2 h compared to gavage of vehicle control in both NGY2R + (−0.267 ± 0.051 g, p < 0.0001, n = 13) and NG Y2R KD mice (−0.221 ± 0.064 g, p = 0.002, n = 12), in accord with mechanical distension and the release of other anorectic gut hormones besides PYY reducing appetite. However, this suppression of food intake was significantly attenuated in the NG Y2R KD compared to NG Y2R + mice (p = 0.044 at 2hr post gavage) (Figure 2C).

Previous studies have suggested that PYY_3-36 inhibits gastric emptying [18], but the pathway mediating this effect is unknown. To investigate whether delayed gastric emptying might play a role in vagal Y2R-mediated feeding suppression, we performed an acetaminophen absorption test in a separate cohort of NG Y2R KD and control-injected NG Y2R + mice. This revealed a significantly higher peak of blood acetaminophen at 15 min in the NG Y2R KD group compared to control injected animals, suggesting that reduced vagal Y2R signalling results in more rapid gastric emptying (Figure 2D). A trend towards an increased area under the curve (AUC) for glucose (278.4 vs 409.9 mmol/L x time; p = 0.1275) in an oral glucose tolerance test (OGTT) was also observed in the NG Y2R KD group (Supp Figs. 3b and 3c). Together, these results suggest that PYY_3-36 released in response to nutrients in the gut slows gastric emptying by acting on vagal afferents, an effect that may contribute to its anorectic effects.

Finally, we investigated whether the vagal Y2R modulated meal patterning. Control-injected NG Y2R + and NG Y2R KD animals were studied in a Comprehensive Laboratory Animal Monitoring Systems (CLAMS) system for 10 days, 10–12 weeks after successful AAV-cre injection. No differences in cumulative food intake, energy expenditure (EE), respiratory exchange ratio (RER) or oxygen consumption were observed in NG Y2R KD animals compared to NG Y2R + controls over the CLAMS period. However, meals were significantly smaller and shorter in NG Y2R KD compared with controls (Figure 2, Figure 3); probability density functions for meal patterning readouts further illustrate highly significant differences (p < 0.0001) in average meal size and duration (Suppl Figs. 3f and 3g), with a trend to shorter intermeal durations, particularly in the dark phase (Figure 3). This was coupled with a non-significant tendency (p = 0.09) for more meals in the KD group over the duration of the CLAMS experiment. There was also a trend towards increased ambulatory activity in the dark phase in NG Y2R KD (Suppl Figs. 3d and e). These data suggest a physiological role for vagal Y2R on meal patterning, even in the absence of a large long-term overall effect on energy balance.

Figure 3.

Expanded CLAMS datasets, day by day, for each individual animal. Results are given for the full day and then split into light and dark periods (12 h each). Grey bars represent the control animals and coloured animals the knockdown animals (NGY2RKD). Red panels show data for meal weights, meal durations in yellow and intermeal duration in orange.

The contribution of gut hormones in the regulation of short-term food intake and longer-term energy homeostasis is well established, and the potential for targeting these hormone systems to promote weight loss has been realised with the licencing of GLP-1 analogues to treat obesity [19]. However, the pathways involved in gut hormone signalling, and the physiological effects of individual hormones are unclear. In particular, the complex role of the vagus nerve in the gut-brain axis is just beginning to be understood. Enteroendocrine cells (EECs) in the gastrointestinal tract postprandially secrete satiety-inducing gut peptides, the targets of which are thought to include vagal afferents that express receptors for these hormones [20]. Gut hormones may diffuse through the lamina propria and interact with vagal afferents terminals, or neuropods on enteroendocrine cells themselves may directly synapse with the vagus nerve [21].

PYY and GLP-1 are co-secreted from L-type EECs and both modulate food intake in humans [22,23]. Our results add to the body of literature suggesting that individual gut hormones act, at least in part, through the vagus nerve to modulate feeding behaviours [24,25]. A recent study suggests that PYY from colonic EECs is responsible for the appetite suppression observed following the activation of these cells [26]. The current study suggests that physiologically released PYY_3-36 requires intact vagal signalling to mediate its effects, but that exogenous administration of higher, pharmacological doses may bypass this pathway. Here we demonstrate that the ability of acute administration of a low dose of PYY_3-36 to reduce food intake over the subsequent 2 h is abrogated when Y2R expression is knocked down in vagal afferents. Although the low dose used here has been shown to result in circulating levels that reflect those observed in the postprandial state [4], the peripheral delivery of this dose may not fully recapitulate local levels experienced by the vagal neuronal terminals. However, the fact that the satiating effects of a mixed meal were also abrogated with vagal afferent Y2R knockdown supports the assertion that the effects of endogenous PYY release on food intake are mediated via the vagus. In contrast, the more pronounced food-intake inhibition of high-dose PYY was equally effective whether or not vagal Y2R had been deleted, which may reflect direct hypothalamic neuronal activation at this dose [27].

In very close agreement with other reports of vagal afferent receptor profiles [28,29] we find that 25–30% of vagal afferent neurones express the Y2R on our RNA scope experiments and 27% of vagal afferent neurones were responsive to PYY in calcium imaging experiments. Mice congenitally lacking the Y2R in their sensory afferent nerves as well as those with adult knockdown of the receptor both demonstrated an abrogated response to low dose PYY but a preserved response to high dose PYY. Congenital Nav1.8 driven loss of Y2R in vagal afferents did not produce an adult body weight or cumulative food intake phenotype compared with littermate controls. This may have been because of the incomplete knockdown of receptors and some redundancy in the system or because of compensatory sequelae. However, mirroring the transient effects of virally-induced knockdown of the hypothalamic Y2R [30], we observed a body-weight phenotype that suggested a positive energy balance in the vagal Y2R KD animals for the first 6 weeks of cre expression, in a postnatal model that achieved similar knockdown efficiency. Since individual food intake was not measured during this period of group housing, it remains to be proven whether this was due to an increase in food intake, as might be expected in a model that blocks the signalling of an anorectic hormone.

By 12 weeks post injection, the body weight phenotype of the vagal Y2R KD animals had reverted to that of the control treated animals, though there was still a clear difference in meal patterning behaviour and gastric emptying. The CLAMS dataset revealed that animals with knocked down Y2R in the afferent vagus engaged in smaller and quicker meals, though with reduced times between feeding bouts, resulting in no overall change in total daily food intake. This may reflect a very specific effect of Y2R signalling on gastric motility, with the caveat that these data were collected when this postnatal knockdown model was no longer exhibiting an overall effect on energy balance and may have at least in part compensated for the receptor loss. There is more detail in the literature regarding GLP-1 receptor activity in the afferent vagus, where it is thought to modulate the sensitivity of mechanosensitive vagal afferent endings [31,32] Thus, loss of the GLP-1 receptor on vagal afferents in rats results in longer and larger meals [24], as might be more expected of a hormone that mediates gastric distension-induced satiety. However, the specific contributions of individual gut hormone pathways are only beginning to be elucidated [33]. Only recently have new techniques in neuroscience and genetics successfully profiled vagal innervation of the gut and established the function of individual neuronal populations [34,35]. Recently, Lovelace et al. showed that Y2R-expressing vagal neurons were responsible for the cardioinhibitory reflex that results in syncope [36]. Further study is required to investigate how the different functions of Y2R neurons in vagus are integrated, and whether different subpopulations of these neurons mediate different effects. For example, our studies were not powered to examine the laterality of Y2R expression or effects on feeding behaviour of the right versus the left vagus, although with modern approaches beginning to unravel the complexities of vagal interoceptive signalling pathways [37], this will be an interesting area for future study.

Further understanding of the phenotype and distribution in vagal afferent endings within the gut would help to distinguish the relative advantages of the Nav1.8 versus the NG injection models. Nav1.8-positive neurons are observed in hormone-sensing mucosal endings as well as in a subgroup of tension-sensing vagal intraganglionic laminar endings, and their anatomical proximity to enteroendocrine cells may vary by neuronal type [33]. It is also worth considering that without cfos or similar data from the brains of these animals, it is possible that rather than direct central effects, the higher doses of PYY were able to activate more of the remaining vagal neurones in a viral knockdown model that is, by nature, not 100% efficient. However, the RNA scope data we provide suggests that the effects on Y2R expression in this model were profound and there are numerous previous studies which confirm the central effects of “pharmacological” doses of PYY [4,10,12]. Recently it has been reported that injection of PYY_3-36 into the DVC (dorsal vagal complex) of the brainstem in rats (where vagal afferents relay signals from innervated organs) had an orexigenic effect and prevented exogenous CCK-induced anorexia [38]. This was suggested to be mediated by DVC Y2R, which is not necessarily incompatible with a peripheral vagal Y2R-mediated PYY gastric emptying effect, further highlighting the complexity of this system.

In summary, these data suggest that Y2R signalling in the afferent vagus nerve plays a physiological role in mediating the anorectic and meal patterning effects of endogenous PYY_3-36. Understanding the various sites and actions of this potent anorexigenic gut hormone may be useful for the development of novel pharmacological approaches to obesity.

CRediT authorship contribution statement

Aldara Martin Alonso: Writing – original draft, Visualization, Project administration, Investigation, Formal analysis. Simon C. Cork: Methodology, Formal analysis. Phyllis Phuah: Methodology, Investigation. Benjamin Hansen: Visualization, Formal analysis. Mariana Norton: Investigation. Sijing Cheng: Investigation. Xiang Xu: Formal analysis. Kinga Suba: Investigation. Yue Ma: Methodology, Investigation. Georgina KC. Dowsett: Methodology, Investigation, Formal analysis. John A. Tadross: Methodology, Investigation, Formal analysis. Brian YH. Lam: Writing – review & editing, Methodology, Investigation. Giles SH. Yeo: Writing – review & editing. Herbert Herzog: Writing – review & editing, Methodology. Stephen R. Bloom: Writing – review & editing. Myrtha Arnold: Methodology. Walter Distaso: Writing – review & editing, Visualization, Formal analysis, Data curation. Kevin G. Murphy: Writing – review & editing, Methodology. Victoria Salem: Writing – review & editing, Writing – original draft, Supervision, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary Figure 1.

Conditioned taste avoidance testing in lean, wildtype C57bl6/J mice (n=8/group) reveals no effect of either low or high doses of intraperitoneally delivered PYY_3-36 in contrast with the known aversive agent lithium chloride.

Supplementary Figure 2: Calcium imaging traces of cultured nodose ganglion neurones loaded with fluo-4 dye. Neurones were cultured from 4 NG (pooled) from control animals (top panel) and 4 NG pooled from NG Y2R KD animals (culled 4 weeks post AAV-cre injection). After 30 seconds baseline imaging (0.5 fr/sec), 1000 nM PYY_3-36 was added and imaging continued for a further 5 minutes prior to the addition of KCl. Only the traces of neurones that responded to both KCl and PYY (ie calcium response >20% above KCl adjusted baseline) are shown here. 21% of control neurones and 7% of Y2RKD neurones were responsive to PYY_3-36.

Supplementary Figure 3.

S3a. Plasma PYY levels in wild-type mice after oral gavage of the same meal bolus used in the feeding studies described in main Figure 2 (n = 4-5). S3b and c. Blood glucose in OGTT and its AUC during 0–120 min in 4-h fasted mice (n = 7-5) in NG Y2R KD and control AAV-GFP NG injected animals (NG Y2R +). Two-way ANOVA followed by Dunnett's test to compare to time 0 (in NG Y2R +, p < 0.01; in NG Y2R KD, p < 0.05). S3d and e. Area under the curve and activity by day/night cycle for ambulatory activity (beam breaks) in NG Y2R KD and AAVGFP injected controls (mean of 3 days, dark phase in grey) in the CLAMS system (n = 6-6). Two-tailed unpaired t test. S3fand g. Probability density function of meal duration (i.e., time taken for feeding bout in seconds) and food intake per feeding session over 7 days in a CLAMS system measuring high resolution food intake. The difference between NG Y2R KD vs NG Y2R + controls is highly significant (p<0.0001) Kolmogorov-Smirnov (K-S) test (n = 6 and 6).

Data availability

Data will be made available on request.